maroffo/claude-forge
An experiment in organizing Claude Code skills. Token-optimized configurations for Go, Python, Rails, and Terraform with patterns extracted from real projects.
ARCHETYPE
Orchestrator
A bench of specialised subagents. The main agent mostly delegates.
Copy this rig
# review before running: this installs third-party code
$ npx degit maroffo/claude-forge/.claude ./rig-claude-forge # inspect, then merge into .claude/ MCP servers are added to Claude Code at local scope; env vars are shown as YOUR_… placeholders — we never store values. Files are fetched with degit into a separate folder so you can review before merging.
$ npx degit maroffo/claude-forge/skills/adr .claude/skills/adr $ npx degit maroffo/claude-forge/skills/android-kotlin .claude/skills/android-kotlin $ npx degit maroffo/claude-forge/skills/apple-swift .claude/skills/apple-swift $ npx degit maroffo/claude-forge/skills/autoresearch-prompt .claude/skills/autoresearch-prompt $ npx degit maroffo/claude-forge/skills/blog-writer .claude/skills/blog-writer $ npx degit maroffo/claude-forge/skills/cloud-infrastructure .claude/skills/cloud-infrastructure $ npx degit maroffo/claude-forge/skills/commit .claude/skills/commit $ npx degit maroffo/claude-forge/skills/cover-image .claude/skills/cover-image $ npx degit maroffo/claude-forge/skills/freeze .claude/skills/freeze $ npx degit maroffo/claude-forge/skills/gemini-review .claude/skills/gemini-review $ npx degit maroffo/claude-forge/skills/golang .claude/skills/golang $ npx degit maroffo/claude-forge/skills/harness-mechanic .claude/skills/harness-mechanic $ npx degit maroffo/claude-forge/skills/harness-trace .claude/skills/harness-trace $ npx degit maroffo/claude-forge/skills/humanizer .claude/skills/humanizer $ npx degit maroffo/claude-forge/skills/ios-debugger .claude/skills/ios-debugger $ npx degit maroffo/claude-forge/skills/knowledge-sync .claude/skills/knowledge-sync $ npx degit maroffo/claude-forge/skills/learning-docs .claude/skills/learning-docs $ npx degit maroffo/claude-forge/skills/learning-loop .claude/skills/learning-loop $ npx degit maroffo/claude-forge/skills/obsidian .claude/skills/obsidian $ npx degit maroffo/claude-forge/skills/orchestrator .claude/skills/orchestrator $ npx degit maroffo/claude-forge/skills/plan-forge .claude/skills/plan-forge $ npx degit maroffo/claude-forge/skills/pr-review .claude/skills/pr-review $ npx degit maroffo/claude-forge/skills/process-clippings .claude/skills/process-clippings $ npx degit maroffo/claude-forge/skills/project-analyzer .claude/skills/project-analyzer $ npx degit maroffo/claude-forge/skills/project-checks .claude/skills/project-checks $ npx degit maroffo/claude-forge/skills/python .claude/skills/python $ npx degit maroffo/claude-forge/skills/rails .claude/skills/rails $ npx degit maroffo/claude-forge/skills/react-nextjs .claude/skills/react-nextjs $ npx degit maroffo/claude-forge/skills/refine-requirements .claude/skills/refine-requirements $ npx degit maroffo/claude-forge/skills/releasing-software .claude/skills/releasing-software $ npx degit maroffo/claude-forge/skills/ruby .claude/skills/ruby $ npx degit maroffo/claude-forge/skills/score .claude/skills/score $ npx degit maroffo/claude-forge/skills/second-opinion .claude/skills/second-opinion $ npx degit maroffo/claude-forge/skills/skill-forge .claude/skills/skill-forge $ npx degit maroffo/claude-forge/skills/source-control .claude/skills/source-control $ npx degit maroffo/claude-forge/skills/terraform .claude/skills/terraform $ npx degit maroffo/claude-forge/skills/test-design-reviewer .claude/skills/test-design-reviewer $ npx degit maroffo/claude-forge/skills/verify-frontend .claude/skills/verify-frontend
$ curl -fsSL --create-dirs -o .claude/agents/AGENT.md https://raw.githubusercontent.com/maroffo/claude-forge/main/agents/architecture-reviewer/AGENT.md $ curl -fsSL --create-dirs -o .claude/agents/AGENT.md https://raw.githubusercontent.com/maroffo/claude-forge/main/agents/database-reviewer/AGENT.md $ curl -fsSL --create-dirs -o .claude/agents/AGENT.md https://raw.githubusercontent.com/maroffo/claude-forge/main/agents/dependency-reviewer/AGENT.md $ curl -fsSL --create-dirs -o .claude/agents/AGENT.md https://raw.githubusercontent.com/maroffo/claude-forge/main/agents/dx-reviewer/AGENT.md $ curl -fsSL --create-dirs -o .claude/agents/AGENT.md https://raw.githubusercontent.com/maroffo/claude-forge/main/agents/harness-mechanic/AGENT.md $ curl -fsSL --create-dirs -o .claude/agents/AGENT.md https://raw.githubusercontent.com/maroffo/claude-forge/main/agents/performance-reviewer/AGENT.md $ curl -fsSL --create-dirs -o .claude/agents/AGENT.md https://raw.githubusercontent.com/maroffo/claude-forge/main/agents/project-analyzer/AGENT.md $ curl -fsSL --create-dirs -o .claude/agents/AGENT.md https://raw.githubusercontent.com/maroffo/claude-forge/main/agents/research-analyst/AGENT.md $ curl -fsSL --create-dirs -o .claude/agents/AGENT.md https://raw.githubusercontent.com/maroffo/claude-forge/main/agents/security-reviewer/AGENT.md $ curl -fsSL --create-dirs -o .claude/agents/AGENT.md https://raw.githubusercontent.com/maroffo/claude-forge/main/agents/software-engineer-pi/AGENT.md $ curl -fsSL --create-dirs -o .claude/agents/AGENT.md https://raw.githubusercontent.com/maroffo/claude-forge/main/agents/software-engineer/AGENT.md $ curl -fsSL --create-dirs -o .claude/agents/AGENT.md https://raw.githubusercontent.com/maroffo/claude-forge/main/agents/tech-writer/AGENT.md $ curl -fsSL --create-dirs -o .claude/agents/AGENT.md https://raw.githubusercontent.com/maroffo/claude-forge/main/agents/test-reviewer/AGENT.md
This rig commits no guardrails. Here is the community baseline instead — the deny/ask rules most often found across all 7,204 rigs:
{
"permissions": {
"deny": [
"Read(./.env)",
"Read(**/.env)",
"Read(~/.ssh/**)",
"Bash(rm -rf *)",
"Read(**/*.pem)",
"Bash(rm -rf /)",
"Bash(git push --force:*)",
"Bash(sudo *)",
"Read(.env)",
"Bash(rm -rf /*)",
"Read(./.env.*)",
"Read(~/.aws/**)",
"Bash(git push --force*)",
"Bash(rm -rf:*)",
"Read(**/*.key)",
"Read(**/.env.*)",
"Bash(sudo:*)",
"Bash(git reset --hard*)",
"Bash(git reset --hard:*)",
"Read(.env.*)"
],
"ask": [
"Bash(git push:*)",
"Bash(git push *)",
"Bash(git commit:*)",
"Bash(rm *)",
"Bash(rm:*)",
"Bash(git rebase *)",
"Bash(wget *)",
"Bash(npm publish:*)",
"Bash(git commit *)",
"Bash(gh pr merge *)"
]
}
} Skills (38)
adrandroid-kotlinapple-swiftautoresearch-promptblog-writercloud-infrastructurecommitcover-imagefreezegemini-reviewgolangharness-mechanicharness-tracehumanizerios-debuggerknowledge-synclearning-docslearning-loopobsidianorchestratorplan-forgepr-reviewprocess-clippingsproject-analyzerproject-checkspythonrailsreact-nextjsrefine-requirementsreleasing-softwarerubyscoresecond-opinionskill-forgesource-controlterraformtest-design-reviewerverify-frontend
Subagents (13)
| architecture-reviewer | Architecture review: SOLID, coupling, cohesion, API design, error handling, patterns |
| database-reviewer | Database review: migration safety, indexes, schema design, query patterns, deadlocks |
| dependency-reviewer | Dependency review: CVEs, licenses, outdated packages, unnecessary deps, bloat |
| dx-reviewer | Developer experience review: ADRs, C4 diagrams, comments, error messages, README, onboarding |
| harness-mechanic | Harness optimization: reads execution traces, identifies systematic failures, proposes rule/skill rewrites with evidence. Launched via /harness-mechanic or when review loop is stuck. |
| performance-reviewer | Performance review: N+1, memory, caching, complexity, allocations, resource pools |
| project-analyzer model: haiku | Analyze codebase structure and create comprehensive CLAUDE.md documentation |
| research-analyst | Pre-plan research: best practices, external repos, docs, prior art before architectural decisions |
| security-reviewer | OWASP-focused security review: injection, auth, secrets, CORS, input validation |
| software-engineer-pi model: haiku | Cost-sensitive implementation executor: drives scripts/pi-exec (pi coding agent + gemini flash, Google-billed) on a scoped brief. Driver only: never edits repo files, never commits, never falls back t |
| software-engineer | Implementation agent: writes code from plans, fixes reviewer findings, scoped to specific workstreams |
| tech-writer | Technical content creator: blog posts, changelogs, release notes from project context |
| test-reviewer | Test quality review: coverage gaps, flaky patterns, edge cases, test architecture |
Hooks (1)
| event | matcher | runs |
|---|---|---|
| SessionStart | * | $CLAUDE_PROJECT_DIR/hooks/checkpoint-reminder.sh |
Similar rigs
rafagomes/ai-workflow
AI-assisted development workflow toolkit for Claude Code, Cursor, and Codex — skills, agents, and conventions for the full SDLC
Skill Collector 4.0k tok ·
christian-deleon/dotfiles
My personal .dotfiles repository.
Skill Collector 5.2k tok ·
chase0213/dotfiles-claude
—
Orchestrator 1.6k tok ·
MiguelAxcar/ai-rpi-protocol
Repo-native protocol for AI-assisted coding that enforces a simple discipline: research first, plan second, code last. Drop it into any repository to reduce wrong implementations, cut rewrite cycles, and improve decisions earlier in the wor
Orchestrator 2.3k tok ·